Methods and apparatus for real-time characterization of batteries with a reference electrode

Abstract

The disclosed battery system comprises a three-electrode metal-ion battery configured with voltage meters connected between anode and cathode, between anode and a reference electrode, and between cathode and the reference electrode; a current source connecting the anode and cathode; and a programmable computer. The system is configured to control the current source to drive the battery with a current cycling profile, and to measure current signals between anode and cathode, and voltage signals derived from the voltage meters. An impulse response is then calculated for each of the anode and cathode, to dynamically estimate open-circuit potential and impedance of each of the anode and cathode. Battery aging, battery capacity fading, and other diagnostics are provided in real time. This invention can characterize each individual electrode of a battery, even when the battery is cycling away from equilibrium states, which is important for electric vehicles.

What is claimed is: 
     
       1. A method of characterizing a metal-ion battery in real time, said method comprising:
 (a) providing or obtaining a battery with a first electrode, a second electrode, and a reference electrode; 
 (b) conducting at least two of the following substeps: (b)(i) providing a first voltage meter connected between said first electrode and said second electrode; (b)(ii) providing a second voltage meter connected between said first electrode and said reference electrode; and/or (b)(iii) providing a third voltage meter connected between said second electrode and said reference electrode; 
 (c) driving said battery, using a current source connecting said first and second electrodes, with a current cycling profile; 
 (d) measuring, in real time, current signals between said first and second electrodes and at least two voltage signals derived from said first, second, and/or third voltage meters in substeps (b)(i), (b)(ii), and/or (b)(iii), respectively; and 
 (e) dynamically characterizing open-circuit potential and impedance of each of said first and second electrodes based on said current signals and said voltage signals. 
 
     
     
       2. The method of  claim 1 , wherein said metal-ion battery is a lithium-ion battery. 
     
     
       3. The method of  claim 1 , wherein all of substeps (b)(i), (b)(ii), and (b)(iii) are conducted. 
     
     
       4. The method of  claim 3 , said method comprising repositioning one of said voltage meters for use as another one of said voltage meters. 
     
     
       5. The method of  claim 1 , wherein said impulse response in step (e) is calculated using a recursive technique. 
     
     
       6. The method of  claim 5 , where said recursive technique comprises constructing an executable mathematical model of said system operable to estimate said impulse response, said mathematical model comprising an equation summing a plurality of sensed data signals including measured current passing said system at a time-certain, and measured voltage of each of said electrodes; updating said impulse response via a recursive least-squares equation based on said sensed data signals at said time-certain, and results determined at a preceding time-certain; and calculating the Fourier transform of said impulse response to obtain an impedance spectrum for each of said electrodes. 
     
     
       7. The method of  claim 1 , wherein said impulse response in step (e) is calculated using a matrix-based technique. 
     
     
       8. The method of  claim 7 , wherein said matrix-based technique comprises the substeps of:
 (e)(i) initializing a state vector, including open-circuit voltage and impulse response of a selected electrode with a finite time sequence; 
 (e)(ii) initializing the covariance matrix with a square matrix; 
 (e)(iii) sensing current and voltage signals of said selected electrode; 
 (e)(iv) constructing input vectors with current signals in said time sequence; 
 (e)(v) constructing output vectors with voltage signals in said time sequence; 
 (e)(vi) computing a signal difference by subtracting the inner product of said state vector and said input vector from said output vector; 
 (e)(vii) updating said covariance matrix; 
 (e)(viii) calculating a gain vector by transforming said input vector with said updated covariance matrix; 
 (e)(ix) updating said state vector and reading out said open-circuit voltage from the first element of said state vector; and 
 (e)(x) updating said open-circuit voltage by repeating steps (iii)-(ix) in a next time step. 
 
     
     
       9. A battery system comprising a three-electrode metal-ion battery configured with at least two voltage meters selected from a first voltage meter connected between a first electrode and a second electrode, a second voltage meter connected between said first electrode and a reference electrode, and/or a third voltage meter connected between said second electrode and said reference electrode; a current source connecting said first and second electrodes; and a computer disposed in communication with said battery, said computer programmed using non-transitory memory with executable code for executing the steps of:
 (a) controlling said current source to drive said battery with a current cycling profile; 
 (b) measuring current signals between said first and second electrodes, and at least two voltage signals derived from said first, second, and/or third voltage meters; and 
 (c) calculating an impulse response of each of said first and second electrodes, from said current signals and said voltage signals, to dynamically estimate open-circuit potential and impedance of each of said first and second electrodes. 
 
     
     
       10. The battery system of  claim 9 , wherein said metal-ion battery is a lithium-ion battery. 
     
     
       11. The battery system of  claim 9 , wherein each of said first, second, and third voltage meters is present in said battery system. 
     
     
       12. The battery system of  claim 9 , wherein said impulse response in step (c) is calculated using a recursive technique. 
     
     
       13. The battery system of  claim 12 , where said recursive technique comprises constructing an executable mathematical model of said system operable to estimate said impulse response, said mathematical model comprising an equation summing a plurality of sensed data signals including measured current passing said system at a time-certain, and measured voltage of each of said electrodes; updating said impulse response via a recursive least-squares equation based on said sensed data signals at said time-certain, and results determined at a preceding time-certain; and calculating the Fourier transform of said impulse response to obtain an impedance spectrum for each of said electrodes. 
     
     
       14. The battery system of  claim 9 , wherein said impulse response in step (c) is calculated using a matrix-based technique. 
     
     
       15. The battery system of  claim 14 , wherein said matrix-based technique comprises the substeps of:
 (i) initializing a state vector, including open-circuit voltage and impulse response of a selected electrode with a finite time sequence; 
 (ii) initializing the covariance matrix with a square matrix; 
 (iii) sensing current and voltage signals of said selected electrode; 
 (iv) constructing input vectors with current signals in said time sequence; 
 (v) constructing output vectors with voltage signals in said time sequence; 
 (vi) computing a signal difference by subtracting the inner product of said state vector and said input vector from said output vector; 
 (vii) updating said covariance matrix; 
 (viii) calculating a gain vector by transforming said input vector with said updated covariance matrix; 
 (ix) updating said state vector and reading out said open-circuit voltage from the first element of said state vector; and 
 (x) updating said open-circuit voltage by repeating steps (iii)-(ix) in a next time step. 
 
     
     
       16. The battery system of  claim 9 , wherein said reference electrode is not spatially between said first and second electrodes. 
     
     
       17. The battery system of  claim 16 , wherein said first electrode is disposed adjacent to a first current collector, wherein said first electrode supplies or accepts selected battery metal ions; said second electrode, with polarity opposite of said first electrode, is disposed adjacent to a second current collector, wherein said second electrode supplies or accepts said metal ions, and wherein said second current collector is porous and permeable to said metal ions; said reference electrode is disposed adjacent to a third current collector, wherein said reference electrode contains said metal ions; wherein a first separator is interposed between said first electrode and said second electrode, to electronically isolate said first electrode from said second electrode; and wherein a second separator is interposed between said second current collector and said reference electrode, to electronically isolate said second electrode from said reference electrode. 
     
     
       18. An apparatus for characterizing a three-electrode metal-ion battery in real time, said apparatus comprising:
 at least two voltage meters selected from first, second, and third voltage meters, wherein said first voltage meter is connectable between a first electrode and a second electrode of a selected battery, said second voltage meter is connectable between said first electrode and a reference electrode of said battery, and said third voltage meter is connectable between said second electrode and said reference electrode; 
 a computer programmed using non-transitory memory with executable code for executing the steps of: 
 (a) controlling a current source to drive said battery with a current cycling profile; 
 (b) measuring current signals between said first and second electrodes, and at least two voltage signals derived from said first, second, and/or third voltage meters; and 
 (c) calculating an impulse response of each of said first and second electrodes, from said current signals and said voltage signals, to dynamically estimate open-circuit potential and impedance of each of said first and second electrodes. 
 
     
     
       19. The apparatus of  claim 18 , wherein each of said first, second, and third voltage meters is present in said apparatus. 
     
     
       20. The apparatus of  claim 18 , wherein said apparatus is linked in operable communication with said battery, and wherein at least two of said two voltage meters are connected between electrodes. 
     
     
       21. The apparatus of  claim 20 , wherein said battery is a lithium-ion battery. 
     
     
       22. The apparatus of  claim 18 , wherein said impulse response in step (c) is calculated using a recursive technique. 
     
     
       23. The apparatus of  claim 22 , where said recursive technique comprises constructing an executable mathematical model of said system operable to estimate said impulse response, said mathematical model comprising an equation summing a plurality of sensed data signals including measured current passing said system at a time-certain, and measured voltage of each of said electrodes; updating said impulse response via a recursive least-squares equation based on said sensed data signals at said time-certain, and results determined at a preceding time-certain; and calculating the Fourier transform of said impulse response to obtain an impedance spectrum for each of said electrodes. 
     
     
       24. The apparatus of  claim 18 , wherein said impulse response in step (c) is calculated using a matrix-based technique. 
     
     
       25. The apparatus of  claim 24 , wherein said matrix-based technique comprises the substeps of:
 (i) initializing a state vector, including open-circuit voltage and impulse response of a selected electrode with a finite time sequence; 
 (ii) initializing the covariance matrix with a square matrix; 
 (iii) sensing current and voltage signals of said selected electrode; 
 (iv) constructing input vectors with current signals in said time sequence; 
 (v) constructing output vectors with voltage signals in said time sequence; 
 (vi) computing a signal difference by subtracting the inner product of said state vector and said input vector from said output vector; 
 (vii) updating said covariance matrix; 
 (viii) calculating a gain vector by transforming said input vector with said updated covariance matrix; 
 (ix) updating said state vector and reading out said open-circuit voltage from the first element of said state vector; and 
 (x) updating said open-circuit voltage by repeating steps (iii)-(ix) in a next time step.